Dose-related effects of flavanol-rich cocoa on blood pressure

Kade Davison1,2, Narelle M Berry1, Gary Misan3,4, Alison M Coates1,3, Jonathan D Buckley1,3 and

Peter RC Howe1,3

1 Nutritional Physiology Research Centre, School of Health Sciences, University of South Australia,

South Australia 5001

2 School of Molecular & Biomedical Sciences, University of Adelaide, South Australia 5005.

3 Australian Technology Network Centre for Metabolic Fitness, University of South Australia,

South Australia 5001.

4 Spencer Gulf Rural Health School, University of South Australia and University of Adelaide,

Whyalla, South Australia 5608.

Short Title: Cocoa and blood pressure

This work was supported by Mars Inc.

Address for correspondence: Prof Peter RC Howe

School of Health Sciences

University of South Australia

GPO Box 2471, Adelaide, SA 5001

Ph +618 8302 1200

CORE Metadata, citation and similar papers at core.ac.uk

Provided by University of East Anglia digital repository

2

Abstract

Background

Consumption of flavanol containing cocoa products has been shown to lower blood pressure (BP),

but the minimum dose required to reduce BP is not known. This study aimed to examine the effect

of three different doses of cocoa flavanols (CF) on 24-hour mean arterial blood pressure.

Method

Twenty four hour ambulatory blood pressure (24-ABP) monitoring was performed in 32 men and

20 postmenopausal women with untreated mild hypertension (seated clinic BP >130/85 and

<160/100 mmHg). Participants were randomised and instructed to consume daily a reconstituted

cocoa beverage containing 33, 372, 712 or 1052 mg/d of CF for 6 weeks in a double-blind, parallel

comparison. Seated clinic BP and 24 hr ABP were measured at 0, 3 and 6 weeks.

Results

Seated clinic BP did not change during the study period. There were significant reductions in 24-

hour systolic (5.3 ± 5.1 mmHg; p = 0.001), diastolic (3 ± 3.2 mmHg; p = 0.002) and mean arterial

blood pressure (3.8 ± 3.2 mmHg; p = 0.0004) at the 1052 mg/d CF only. No reduction in BP was

seen at any other dose.

Conclusion

No evidence of dose response was seen in this experiment. The highest dose of 1052 mg CF/d was

found to significantly lower BP. These results support previous evidence for CF to lower BP,

however more research is needed to establish the most effective dose and food matrix.

3

Introduction

Several intervention studies have reported beneficial effects of consuming cocoa containing

products on blood pressure (BP) [1-5] and a recent meta-analysis confirmed this effect by

demonstrating that the consumption of cocoa-rich foods for seven or more days has the capacity to

lower resting BP in normotensive and mildly hypertensive adults [6]. This effect of cocoa products

has been widely attributed to its content of flavanols and procyanidins (oligomeric and polymeric

flavanols), which for the purpose of this paper will be included in the term “cocoa flavanols” (CF)

[7, 8].

The recent meta-analysis [6] included five studies which compared the antihypertensive effects of

CF-rich chocolate with chocolate containing few or no CF. The amount of chocolate consumed

daily ranged from 46 to 105g and provided intakes of CF ranging from 246 to 500 mg/day. There

was a mean BP reduction of 4.7 / 2.8 mmHg (systolic/diastolic). In a subsequent study that was not

included in the meta-analysis, Taubert et al [9] showed that consuming as little as 30 mg of CF per

day in 6g of dark chocolate for 18 weeks was sufficient to lower systolic blood pressure (SBP) by

2.9 mmHg and diastolic blood pressure (DBP) by 1.9 mmHg. A more recent meta analysis by

Hooper et al [10] also reported a net effect in favour in cocoa in the order of 6mmHg systolic and

3mmHg diastolic. Although the potential antihypertensive effect of chocolate consumption is

interesting, the high sugar and fat content of chocolate undermines its potential as a health

enhancing food [11, 12].

Concern about delivering CF in a high sugar and high fat chocolate food matrix has led to

investigations of the potential antihypertensive and other cardiometabolic health benefits of

relatively low fat drinks enriched with CF. In a study which examined the acute effect of consuming

a flavanol-rich cocoa beverage, Schroeter et al [13] demonstrated an increase in flow mediated

dilatation of the brachial artery (FMD). Four studies have been conducted examining the longer-

term effects of consuming flavanol-rich cocoa beverages on cardiovascular health in various

populations. In the first of these daily consumption of 446mg of cocoa flavanols by post-

4

menopausal women for 6 weeks produced a significant improvement in arterial function but no

change in BP compared to placebo [14]. In another study individuals with type 2 diabetes took

963mg /day for a period of 30 days and also experienced improvements in endothelial function but

not BP [15]. In a third study in which overweight or obese individuals consumed 902mg/day of

flavanols for 12 weeks, there were improvements in endothelial function with modest yet significant

reductions in BP (MAP reduced by 1.2 mmHg) [16]. The fourth study provided approximately

900mg/day of CF to individuals with essential hypertension for 2 weeks [17]. Despite the similarity

of this protocol to that by which Grassi et al [4] showed BP reductions with chocolate, and despite

providing almost double the daily dose of CF, there was no change in BP although there was some

improvement in endothelial function. The lack of BP change in three of these studies and the

relatively small change in the other following chronic consumption of large doses of CF are

inconsistent with the results seen in the meta-analysis [6] and subsequent study [9] with flavanol

rich chocolate consumption. These findings suggest that the dose of CF alone does not determine

the change in BP, however a number of differences in the design of these studies makes direct

comparisons difficult. The studies that have been published to date examined different patient

populations (healthy and hypertensive adults in chocolate based protocols and

hypercholesterolaemic, overweight/obese and type 2 diabetic in non-chocolate protocols), and the

type 2 diabetes study was confounded by participant medications. A further complication with

comparison is a lack of consistency in the technique of BP assessment. Most studies have used

seated clinic BP assessments with only three studies to date utilising the preferred technique of

ambulatory blood pressure (ABP) monitoring [3, 4, 18]. These three studies used the same protocol

of a 15 day crossover study with dark chocolate (delivering approximately 500mg of CF) or white

chocolate (CF free) in three separate subject groups including essential hypertensives and have

reported the largest reductions in BP to date . With these factors in mind, a single study evaluating

the potential BP lowering effects of CF in a homogenous patient population using the preferred

method of ABP assessment and delivering the CF in a consistent food matrix was warranted.

5

Therefore, the primary aim of the present study was to determine the dose-response effect of CF

delivered using a flavanol-rich cocoa beverage on 24-hour ambulatory mean arterial pressure

(MAP) in an untreated borderline/mild hypertensive population. Secondary outcomes include clinic

BP assessment and additional measures of 24-hour ABP.

6

Method

Participants

Male and post-menopausal female adults with high-normal BP or untreated mild hypertension (SBP

130-160 mmHg or DBP 85-100 mmHg) were recruited by public advertisement. Participants were

excluded if they had a known diagnosis of cardiovascular disease (or a history of cardiovascular or

cerebrovascular incidents); Diabetes (Type 1 or Type 2) or were taking prescribed anti-diabetic

medication; renal failure; were taking BP lowering medication or supplements that may influence

BP (i.e. fish oil, liquorice, polyphenols) in the preceding 3 months, or were likely to do so during

the study period; an intolerance to alkaloids (caffeine, theobromine) or dairy; were currently

smoking or using nicotine replacement therapy; or had any other medical condition which may

have influenced the outcome of the study. The study was approved by the Human Research Ethics

Committee of the University of South Australia. Each participant provided written and informed

consent prior to participation. Recruitment was carried out between February and August of 2007

and the intervention was progressively conducted from April to October of 2007.

Eligibility screening

Potential participants were initially screened for eligibility by completing a health and lifestyle

questionnaire and undertaking a seated clinic BP assessment. Participants with a seated BP between

SBP 130-160 mmHg or DBP 85-100 mmHg were enrolled in the study. Enrolled participants

returned to the clinic approximately one week later to have their BP measured under the same

conditions as the initial screening to confirm the presence of high-normal BP or mild hypertension.

If they requalified (i.e. SBP 130-160 mmHg or DBP 85-100 mmHg), they commenced the study

protocol. If their BP fell outside these limits they were invited to return for an additional qualifying

BP assessment. Participants with screening BP consistently above thresholds for BP (i.e. diastolic >

100 mmHg; systolic > 160 mmHg) were referred to their GP and not entered into the study.

7

Protocol outline

Participants were block-matched by the minimisation convention [19] on BP, gender, age and BMI

into four treatment groups which were randomised to consume reconstituted cocoa beverages

containing 33, 372, 712 or 1052 mg/d of CF for 6 weeks in a double-blind, parallel comparison.

Randomisation of groups was undertaken independently of group minimisation procedure by

separate staff members of the research centre not otherwise involved with the trial. Trial

investigators remained blinded to treatment allocation until after the completion of data analysis.

Participants then had body weight assessed and underwent seated clinic BP and 24 hour ABP

assessments at baseline and were then required to consume a cocoa drink daily for the next 6 weeks.

Seated clinic BP and 24 hour ABP assessments were repeated after 3 and 6 weeks (Table 2).

Cocoa supplements

The cocoa drinks were prepared by participants mixing the contents of three cocoa beverage sachets

with 300 mL of water each morning and drinking 30 minutes prior to breakfast. The cocoa flavanol

composition of the cocoa beverages are provided in Table 1. Sachets (dry powder – 58g net) were

matched for micro and macro nutrients including total energy (mean ~ 910 kJ); saturated fat (mean

~ 1.4g); sugar (approx. mean ~ 17.5g); caffeine (mean ~ 43mg); and theobromine (mean ~ 430mg).

On all testing days (at 3 and six weeks), volunteers were instructed not to consume their cocoa

beverage until after testing was completed to eliminate any acute effects of cocoa consumption. All

empty sachets were returned at 3 and 6 weeks to monitor compliance. Participants were instructed

to report any signs or symptoms of ill health immediately to the researchers and any action that was

taken (e.g. medication) to identify any potential adverse effect of the intervention. In addition to this

participants were specifically asked at each visit whether any signs or symptoms of ill health had

been experienced. All reported events were documented by the researchers in each participant’s

case report form.

8

Blood pressure monitoring

Seated clinic BP was assessed after sitting quietly for at least 5 min using an automated

oscillometric BP monitor (HDI/Pulsewave CR-2000 Cardiovascular Profiler, Hypertension

Diagnostics Inc, Eagan, MN) in accordance with the procedures outlined by the Joint National

Committee on the Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (VII):

US Dept of Health and Human Services [20]. Four consecutive BP readings were taken at 1 min

intervals by a single observer. The first reading was discarded and an average of the remaining

measurements was taken to determine eligibility for study entry.

For assessment of 24 hour Ambulatory Blood Pressure (ABP), volunteers were fitted with a

SpaceLabs ambulatory BP monitor (Model 90217, SpaceLabs Medical, Florida, USA) for 24 hours.

An appropriately sized cuff was placed firmly around the upper non-dominant arm, centred over the

brachial artery, with the monitor worn on a waist strap; ABP measurements were recorded at 15

minute intervals excepting from 11 pm – 7am (30 minute intervals). The cuff and monitor was only

removed briefly for bathing and remained in place at all other times during the 24 hour recording

period. Assessments at baseline, 3 and 6 weeks were carried out on the same day of the week using

the same monitor, cuff size and arm. Participants were required to maintain an activity diary during

each 24 hour ABP period to enable BP values to be related to activities. 24 hour ABP recordings

were used to determine daytime (7am – 11pm), night-time (11pm -7am) and 24-hour averages for

MAP, SBP, DBP and heart rate (HR).

Diet and lifestyle requirements

Volunteers were asked to maintain their normal physical activity patterns and were provided with a

list of dietary exclusions to ensure a low-flavanol diet for 1 week prior to and during the study

9

period. Given the energy content of the beverages (~900 kJ), volunteers were advised on

appropriate diet substitutions to avoid increasing total energy intake. To monitor compliance with

this request body weight was measured at each clinic visit.

Statistical analysis

It was estimated that a sample size of 12 subjects per arm would have at least 80% power to detect a

statistically significant difference in reduction in MAP between the four treatment groups (allowing

for multiple comparisons) using an ANOVA at the 0.05 significance level. The estimates were

based upon the assumption of a linear dose response effect achieving a 5mmHg change in 24 hour

MAP with the highest dose and a standard deviation of 4mmHg as indicated by unpublished pilot

data. These estimates are supported by the previously reported ABP results with CF consumption

(3, 4, 18).

Baseline characteristics were compared between groups by one-way analysis of variance

(ANOVA). Comparison of changes in ABP between treatment groups (i.e. CF dose) across time

(i.e. 3 and 6 week assessments) was made using repeated measures analysis of covariance

(ANCOVA) using baseline ABP as the covariate. Where ANCOVA showed a statistically

significant main effect, post-hoc comparisons (Tukey’s HSD) were conducted to identify

differences between means. To optimise the analysis of dose effects on ABP, a nested ANOVA

design was used to examine changes in ABP from baseline by 3 and 6 weeks with time (i.e. weeks 3

and 6) nested in dose. Comparison of changes in HR between treatment groups across time was

made using repeated measures ANOVA. Statistical significance was set at p<0.05. All data are

presented as mean ± SEM.

10

Results

Participant flow is provided in Figure 1. In total, 53 volunteers completed the study. One volunteer

was excluded due to mild persistent (>3 day) gastric symptoms that may have been related to the

test beverage. No other adverse events related to the study were reported. Five volunteers withdrew

during the study due to personal circumstances unrelated to the study. One volunteer was excluded

from the analysis due to non-compliance to the study protocol, leaving 52 volunteers, or n=12-14

volunteers per arm. Trial withdrawals and exclusions did slightly impact upon the success of the

minimisation process on the matching of groups for the stated baseline parameters as can be seen in

Table 1. However, no significant between group differences were seen for the blocking parameters

including age (p=0.45), BMI (p=0.28), and BP (Systolic p=0.86; Diastolic p=0.45). Compliance

with consumption of cocoa supplements was greater than 98% in all groups. Body mass did not

change during the intervention (p = 0.48 for treatment x time).

SBP, DBP, MAP or HR for 24-hour, night, day, and seated clinic measurements for each treatment

group are provided in Tables 2-5. There were no dose x time interactions for seated clinic BP, but a

significant time effect for SBP (p=0.02) and MAP p=0.01) was evident. However, there were

significant dose x time effects for 24 hour MAP (p=0.047) and overnight HR (p = 0.041) assessed

by ABP monitoring. ANCOVA of the change from baseline (week 0) to weeks 3 and 6 (with

baseline values as a covariate) revealed a significant effect of dose on 24-hour ambulatory SBP

(p=0.019), DBP (p=0.017) and MAP (p=0.008) and night ambulatory SBP (p=0.0006), DBP

(p=0.005) and HR (p=0.043).

Nested analysis (time nested in dose) revealed significant dose effects for 24 hour MAP (p=0.0004),

SBP (0.001) and DBP (p=0.002); night time MAP (p=0.01; Fig 3) and SBP (p=0.003) and day time

MAP (p=0.02; Fig 3) and SBP (p=0.02). Post hoc analysis showed that for the 1052 mg flavanol

11

dose, the reductions in for 24 hour MAP (p<0.001), SBP (p<0.02) and DBP (p<0.04) were

significantly different from all other doses (refer Figure 2). There were no significant effects on HR.

Figure 3 displays a consistent reduction in day and night ABP to that seen over the full 24 hours,

with significant effects observed at a CF level of 1052 mg/d.

12

Discussion

The principal finding of the current study was that the regular consumption of flavanol-rich

reconstituted cocoa beverages reduced BP in untreated patients with borderline/mild hypertension.

This antihypertensive effect was only evident at the highest dose of 1052 mg flavanols/day, with

doses up to and including 712 mg/day failing to provide any significant reduction in BP in this

study. This is the first study to directly compare the efficacy of differing dose levels of isolated CF.

The doses used were based upon the range of CF levels that have been previously reported to lower

BP. These findings suggest that BP can be lowered through the regular consumption of flavanol-

rich cocoa beverages, though higher levels of CF may be required than has been observed with

consumption of chocolate based CF rich products. Although this study provided the highest dose of

CF to date the lack of effect on BP with lower doses is supported by previous studies using a non-

chocolate food matrix. Whilst there was no evidence of a dose response effect over the doses tested

in this trial, it may be the case that the highest dose in this study is on the lower end of a dose

response curve with this type of CF product. Additionally, the maximum potential effect on BP is

yet to be determined.

Four previous studies have investigated the longer-term effects of consuming flavanol-rich cocoa

beverages on BP [14-16] using beverages providing daily flavanol doses ranging from 446 – 964

mg. Of these studies only one demonstrated a reduction in BP and this was a study which provided

one of the highest doses of flavanols (902 mg/day). The finding of this study is in agreement with

the outcome of the current study which demonstrated that a high daily flavanol dose is required in

order to achieve a reduction in BP. However, these findings are in contrast with those which have

examined the BP lowering effects of cocoa flavanols delivered in a chocolate food matrix where

antihypertensive effects have been demonstrated with flavanol doses which were reported to be

significantly lower than the levels used in the studies with cocoa beverages. A recent meta-analysis

[9] which examined the antihypertensive effects of flavanol-rich chocolate found that daily

13

consumption of chocolate providing relatively modest doses of cocoa flavanols and procyanidins

ranging from 246 mg to 500 mg provided significant reductions in BP. Subsequently, Taubert et al

[9] showed that consuming as little as 30 mg of cocoa polyphenols per day in dark chocolate for 18

weeks was sufficient to lower systolic blood pressure (SBP) by 2.9 mmHg and diastolic blood

pressure (DBP) by 1.9 mmHg. This latter flavanol dose is similar to the lowest flavanol dose

provided in the current study where no effect on BP was observed. Thus, it appears there may be a

discrepancy in the effective dose of flavanols when delivered in a chocolate matrix compared to a

beverage mix. A previous bioavailability study showed that there was no difference in the acute

increases in plasma or urinary concentrations of flavanols following consumption of chocolate or

dry cocoa containing the same quantities of flavanols [21], and similar bioavailability of cocoa

flavanols from chocolate and other sources is also supported by comparisons from other studies [1,

13, 21]. There has been some debate over the potential for the addition of milk protein to cocoa

products to reduce flavanol bioavailability [22] and this would potentially impact on efficacy when

consumed in milk drinks. However, other recent studies have not supported any effect of milk on

reducing flavanol bioavailability [22-24].

A noteworthy protocol consideration when comparing the studies using flavanol-rich chocolate and

flavanol-rich beverages are the degree of placebo control and blinding. Studies using chocolate have

typically used a flavanol-poor white chocolate placebo which limits subject blinding and is not

matched for all potentially bioactive nutrients including methylxanthines (3-4, 9). The use of

flavanol-rich cocoa beverages allows close matching of placebo beverage nutrient composition as

well as close matching for colour and flavour to facilitate a double blinded protocol (13-16).

Muniyuppa et al (17) propose that the discrepancy between the results seen with dark chocolate and

flavanol-rich cocoa beverages may be entirely due to these factors. The robust placebo control and

blinding used in the present study combined with ABP monitoring provides a superior assessment

of the anti-hypertensive potential of CF to those conducted previously.

14

The mechanism by which CF reduce BP has been largely attributed to their capacity to improve

endothelial dilatory function [8, 25]. A number of studies have demonstrated that the regular intake

of cocoa flavanols can improve endothelium-mediated vascular dilatory function, with these

improvements being associated with an increased bioavailability of nitric oxide (NO) [2, 4, 25, 26,

28]. Recently Taubert et al [9] established a direct relationship which suggested that the BP

lowering effects of cocoa flavanols were mediated by improved NO availability when they found a

statistically significant correlation between the magnitude of reduction in BP and the magnitude of

increase in S-nitrosoglutathione (a marker of NO availability) following 18weeks of consuming

chocolate containing cocoa flavanols. However, some studies have demonstrated improvements in

vascular dilatory function independently of changes in BP, suggesting that other mechanisms might

also contribute to the antihypertensive effect. Two recent studies [14, 15] reported a sustained (non-

acute) improvement in markers of endothelium-mediated dilatory function without concurrent

changes to BP, while a third study [16] conducted in our laboratory showed improvement in both

parameters but only a modest reduction was seen in BP while endothelium-mediated dilatory

function improved by approximately 40%.

The lack of a demonstrable effect of cocoa consumption on seated clinic BP in the present study

may also reflect the relative lack of sensitivity of this technique compared with ABP monitoring.

Seated clinic BP reduced with time independently of flavanol dose, reflecting a habituation effect

(regression to the mean) which most likely masked any treatment effect on BP. An habituation

effect was not evident with ABP monitoring. In a previous study, we were able to demonstrate a

statistically significant reduction in BP with clinic assessments [16], but the BP assessment

protocol in that study was different from that used in the current study with clinic BP measured

while participants were supine after a longer pre-assessment rest period. Additionally for the

15

previous study patients with elevated BP were not specifically recruited which may have reduced

the likelihood of a ‘white coat’ hypertensive effect masking any effect of the cocoa supplement.

A key concern with the consumption of bioactive nutrients for health benefits is the question of

potential adverse effects of the other ingredients. In the case of cocoa flavanols this concern is

particularly associated with the high energy and saturated fat content of chocolate. According to

HPLC analysis of flavanol content of various food types, to achieve the effective dose of flavanol

seen in this study (1052mg) would require the daily consumption of 210g of standard dark

chocolate or approximately double this amount of milk chocolate [29]. Given that 210 grams of

dark chocolate provides in the order of 4700Kj and 35g of saturated fat a net health effect may be

questionable [30]. This is in contrast to the 914kJ and 1.4g of saturated fat provided by the

beverage used in this trial. However, depending upon the underlying causes of the previously

discussed discrepancy between the effect of chocolate and flavanal rich cocoa beverages on BP,

this comparison may not prove valid. It should be noted that the intention of this trial was not to

evaluate the use of a flavanol rich beverage as a potential food supplement, but rather to determine

the effect of various doses of cocoa flavanols on BP.

Ideally the lowest dose would deliver 0mg of flavanol; however, this would be not possible with

our cocoa based product without significantly altering the flavour and or consistency of the drink,

thereby compromising the treatment blinding. Although Taubert et al [9] reported an

antihypertensive effect of 30mg CF/day, this has only been seen with a white vs. dark chocolate

protocol and not with a well matched control so interpretation of these data are subject to the

limitations listed above. Moreover others have failed to show effects on BP of considerably higher

doses of isolated CF [14, 15]. Other concerns regarding the 33mg dose results in the present study

relate to the lower age and BMI and greater proportion of males in this group. While care was taken

to avoid this in the blocking process, with the relatively small subject numbers per treatment arm

16

small numbers of participant attrition can disrupt this balance. It could be argued that this may

reduce the potential responsiveness of the control group, however the lack of BP reduction seen at

the two intermediate doses suggests this has not influenced the overall results. Additionally, the

differences seen in BMI and age were not statistically significant and the most important variable

(BP) remained well matched across the four groups. The final limitation was the lack of explicit

assessment of the effectiveness of the blinding protocol. The close matching of the cocoa products

implies successful blinding, however these results could have been strengthened by assessing this.

In summary, this study provides the most robust assessment of the effects of cocoa flavanols on

blood pressure to date. A significant antihypertensive effect of daily cocoa flavanol intake was

observed at the highest dose of 1052mg flavanols/day with no BP lowering effect seen at the lower

doses. This study supports the potential for cocoa flavanols to lower BP, but further research is

required to determine the extent of antihypertensive benefit that can be achieved with different

dietary sources and doses of flavanols in various cardiovascular pathologies.

Acknowlegements

The authors would like to acknowledge the contribution of all the study participants and MARS inc

for the supply of the test product and financial support (by way of an Australian Government Food

Innovation Grant) to conduct study.

17

References

1. Engler, M.B., et al., Flavonoid-rich dark chocolate improves endothelial function and increases plasma epicatechin concentrations in healthy adults. J Am Coll Nutr, 2004. 23(3): p. 197-204.

2. Fraga, C.G., Cocoa, diabetes, and hypertension: should we eat more chocolate? Am J Clin Nutr, 2005. 81(3): p. 541-2.

3. Grassi, D., et al., Short-term administration of dark chocolate is followed by a significant increase in insulin sensitivity and a decrease in blood pressure in healthy persons. Am J Clin Nutr, 2005. 81(3): p. 611-4.

4. Grassi, D., et al., Cocoa reduces blood pressure and insulin resistance and improves endothelium-dependent vasodilation in hypertensives. Hypertension, 2005. 46(2): p. 398-405.

5. Taubert, D., et al., Chocolate and blood pressure in elderly individuals with isolated systolic hypertension. JAMA, 2003. 290(8): p. 1029-30.

6. Taubert, D., R. Roesen, and E. Schomig, Effect of cocoa and tea intake on blood pressure: a meta-analysis. Arch Intern Med, 2007. 167(7): p. 626-34.

7. Heiss, C., et al., Vascular effects of cocoa rich in flavan-3-ols. JAMA, 2003. 290(8): p. 1030-1.

8. Keen, C.L., et al., Cocoa antioxidants and cardiovascular health. Am J Clin Nutr, 2005. 81(1 Suppl): p. 298S-303S.

9. Taubert, D., et al., Effects of low habitual cocoa intake on blood pressure and bioactive nitric oxide: a randomized controlled trial. JAMA, 2007. 298(1): p. 49-60.

10. Hooper, L., et al., Flavonoids, flavonoid-rich foods, and cardiovascular risk: a meta-analysis of randomized controlled trials. Am J Clin Nutr, 2008. 88(1): p. 38-50.

11. Ferri, C., D. Grassi, and G. Grassi, Cocoa beans, endothelial function and aging: an unexpected friendship? J Hypertens, 2006. 24(8): p. 1471-4.

12. Ferri, C. and G. Grassi, Mediterranean diet, cocoa and cardiovascular disease: a sweeter life, a longer life, or both? J Hypertens, 2003. 21(12): p. 2231-4.

13. Schroeter, H., et al., (-)-Epicatechin mediates beneficial effects of flavanol-rich cocoa on vascular function in humans. Proc Natl Acad Sci U S A, 2006. 103(4): p. 1024-9.

14. Wang-Polagruto, J.F., et al., Chronic consumption of flavanol-rich cocoa improves endothelial function and decreases vascular cell adhesion molecule in hypercholesterolemic postmenopausal women. J Cardiovasc Pharmacol, 2006. 47 Suppl 2: p. S177-86; discussion S206-9.

15. Balzer, J., et al., Sustained benefits in vascular function through flavanol-containing cocoa in medicated diabetic patients a double-masked, randomized, controlled trial. J Am Coll Cardiol, 2008. 51(22): p. 2141-9.

16. Davison, K., et al., Effect of cocoa flavanols and exercise on cardiometabolic risk factors in overweight and obese subjects. Int J Obes (Lond), 2008. 32(8): p. 1289-96.

17. Muniyappa, R., et al., Cocoa consumption for 2 wk enhances insulin-mediated vasodilatation without improving blood pressure or insulin resistance in essential hypertension. Am J Clin Nutr, 2008. 88(6): p. 1685-96.

18. Grassi, D., et al., Blood pressure is reduced and insulin sensitivity increased in glucose-intolerant, hypertensive subjects after 15 days of consuming high-polyphenol dark chocolate. J Nutr, 2008. 138(9): p. 1671-6.

19. Altman, D.G. and J.M. Bland, Treatment allocation by minimisation. BMJ, 2005. 330(7495): p. 843.

20. Chobanian, A.V., et al., Seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. Hypertension, 2003. 42(6): p. 1206-52.

18

21. Baba, S., et al., Bioavailability of (-)-epicatechin upon intake of chocolate and cocoa in human volunteers. Free Radic Res, 2000. 33(5): p. 635-41.

22. Serafini, M., et al., Plasma antioxidants from chocolate. Nature, 2003. 424(6952): p. 1013. 23. Schroeter, H., et al., Nutrition: milk and absorption of dietary flavanols. Nature, 2003.

426(6968): p. 787-8; discussion 788. 24. Keogh, J.B., J. McInerney, and P.M. Clifton, The effect of milk protein on the bioavailability

of cocoa polyphenols. J Food Sci, 2007. 72(3): p. S230-3. 25. Heiss, C., et al., Endothelial function, nitric oxide, and cocoa flavanols. J Cardiovasc

Pharmacol, 2006. 47 Suppl 2: p. S128-35; discussion S172-6. 26. Heiss, C., et al., Sustained increase in flow-mediated dilation after daily intake of high-

flavanol cocoa drink over 1 week. J Cardiovasc Pharmacol, 2007. 49(2): p. 74-80. 27. Fisher, N.D., et al., Flavanol-rich cocoa induces nitric-oxide-dependent vasodilation in

healthy humans. J Hypertens, 2003. 21(12): p. 2281-6. 28. Schnorr, O., et al., Cocoa flavanols lower vascular arginase activity in human endothelial

cells in vitro and in erythrocytes in vivo. Arch Biochem Biophys, 2008. 476(2): p. 211-5. 29. Hammerstone, J.F., S.A. Lazarus, and H.H. Schmitz, Procyanidin content and variation in

some commonly consumed foods. J Nutr, 2000. 130(8S Suppl): p. 2086S-92S. 30. Food Standards Australia New Zealand, Nutrient Tables for use in Australia. 2006, Food

Standards Australia New Zealand.

19

Table 1: Flavanol composition (monomers and polymers/oligomers) of daily beverage for each dose group.

Group No. 1 2 3 4Total flavanol dose 33 372 712 1052Epicatechin 0 69 138 208Catechin 12 28 43 58Other flavanols 20 275 530 785(polymeric/oligomeric) Macronutrient Weight (g) 58 58 58 58Energy (kJ) 904.5 907.9 911.2 914.6Total fat (g) 2.6 2.6 2.6 2.6Sat fat (g) 1.4 1.4 1.4 1.4Total CHO (g) 31.0 31.2 31.5 31.8Sugars (g) 17.3 17.4 17.6 17.7Protein (g) 17.1 17.1 17.1 17.1Caffeine (mg) 46.6 44.7 42.7 40.8Theobromine (mg) 402.2 421.6 441.1 460.5

20

Table 2: Baseline characteristics. (mean ± SEM). M/F = ratio male to female; BMI = body mass index.

Group (mg flavanol/day) 33 mg 372 mg 712 mg 1052 mg N 14 12 13 13

Age (years) 53.0 ± 6.7 56.2 ± 14.2 60.2 ± 13.7 56.8 ± 9.7 M/F 10/4 7/5 8/5 7/6

BMI (kg/m2) 33.7 ± 16.8 28.7 ± 5.5 26.9 ± 16.2 27.9 ± 5.0

21

Table 3: Systolic BP (mmHg) for 24 hour, night, day and seated clinic measurement in each

treatment group (mg flavanol/day) at each assessment point (mean ± SD).

Dose (mg)

BP (mmHg) Week 0 Week 3 Week 6 ∆ week 3 ∆ week 6

33

N=14

24hr 133.3 ± 11.6 132.5 ± 14.6 133.9 ± 13.8 -0.8 ± -5.6 0.6 ± 7.1Night 117.1 ± 14.2 116.9 ± 15.3 118.6 ± 10.5 -0.2 ± 6.4 1.5 ± 8.6Day 140.1 ± 11.2 140.0 ± 15.0 141.2 ± 15.7 -0.1 ± -7.1 1.1 ± 10.1

Clinic 145.4 ± 9.7 142.1 ± 10.5 143.3 ± 13.1 -3.3 ± -6.0 -2.1 ± -9.0

372

N=12

24hr 133.1 ± 11.2 132.9 ± 13.3 133.4 ± 13.0 -0.2 ± -5.0 0.3 ± 6.5Night 115.5 ± 12.3 115.6 ± 10.5 115.4 ± 13.7 0.1 ± 6.5 -0.1 ± -8.7Day 141.1 ± 12.3 140.8 ± 15.1 142.5 ± 14.8 -0.3 ± -6.5 1.4 ± 6.9

Clinic 142.9 ± 10.5 142.0 ± 11.9 139.6 ± 9.0 -0.9 ± -9.0 -3.3 ± -7.9

712

N=13

24hr 127.4 ± 7.6 127.9 ± 9.0 128.9 ± 8.3 0.5 ± 6.2 1.5 ± 5.9Night 111.0 ± 9.7 113.1 ± 10.0 114.7 ± 10.7 2.1 ± 8.7 3.7 ± 5.5Day 134.6 ± 7.6 134.4 ± 9.4 135.6 ± 8.0 -0.2 ± -5.5 1.0 ± 6.6

Clinic 143.0 ± 9.0 140.2 ± 10.4 140.0 ± 9.4 -2.8 ± -6.6 -3.0 ± -9.7

1052

N=13

24hr 127.8 ± 9.4 121.0 ± 10.7 124.0 ± 10.7 -6.8 ± -5.9 -3.8 ± -5.5Night 113.5 ± 11.8 106.8 ± 9.7 110.1 ± 7.3 -6.7 ± -8.0 -3.4 ± -8.0Day 133.2 ± 9.4 127.4 ± 12.1 129.8 ± 13.5 -5.8 ± -5.2 -3.4 ± -7.3

Clinic 143.0 ± 8.0 139.8 ± 11.1 138.9 ± 13.5 -3.2 ± -3.1 -4.1 ± -12.1

22

Table 4: Diastolic BP (mmHg) for 24 hour, night, day and seated clinic measurement in each

treatment group (mg flavanol/day) at each assessment point (mean ± SD).

Dose (mg) BP (mmHg) Week 0 Week 3 Week 6 ∆ week 3 ∆ week 6 33

N=14

24hr 82.2 ± 7.1 81.6 ± 7.9 82.8 ± 7.9 -0.6 ± -3.4 0.6 ± 4.1 Night 70.0 ± 8.6 69.7 ± 9.0 71.6 ± 6.4 -0.3 ± -5.6 1.6 ± 4.5 Day 87.2 ± 6.4 86.7 ± 8.2 87.6 ± 9.0 -0.5 ± -4.5 0.4 ± 6.4

Clinic 88.0 ± 6.0 87.7 ± 7.5 88.1 ± 7.9 -0.3 ± -3.7 0.1 ± 4.9

372

N=12

24hr 81.3 ± 8.7 80.6 ± 10.0 81.4 ± 10.0 -0.7 ± -2.8 0.1 ± 3.6 Night 69.3 ± 7.6 69.1 ± 9.0 69.3 ± 8.7 -0.2 ± -4.5 0.0 ± 5.8 Day 87.0 ± 10.0 86.1 ± 11.4 87.2 ± 11.1 -0.9 ± -3.5 0.2 ± 3.6

Clinic 86.4 ± 7.6 85.9 ± 8.0 83.9 ± 9.7 -0.5 ± -5.2 -2.5 ± -5.8

712

N=13

24hr 78.3 ± 9.4 78.2 ± 9.4 79.1 ± 10.5 -0.1 ± -3.2 0.8 ± 2.5 Night 66.9 ± 8.7 67.8 ± 10.5 69.6 ± 11.5 0.9 ± 5.4 2.7 ± 5.4 Day 83.2 ± 10.1 82.6 ± 9.7 83.6 ± 10.5 -0.6 ± -3.2 0.4 ± 3.2

Clinic 85.9 ± 9.4 84.4 ± 11.5 84.2 ± 10.1 -1.5 ± -5.8 -1.7 ± -5.0

1052

N=13

24hr 76.3 ± 9.0 72.5 ± 9.4 74.2 ± 8.7 -3.8 ± -3.6 -2.1 ± -3.2 Night 66.0 ± 8.3 63.3 ± 8.3 65.0 ± 6.9 -2.7 ± -4.7 -1.0 ± -4.3 Day 80.0 ± 9.4 76.5 ± 10.1 78.3 ± 9.4 -3.5 ± -3.2 -1.7 ± -4.7

Clinic 83.2 ± 9.4 81.2 ± 10.8 81.2 ± 8.7 -2.0 ± -6.1 -2.0 ± -3.2

23

Table 5: Mean arterial pressure (mmHg) for 24 hour, night, day and seated clinic measurement in

each treatment group (mg flavanol/day) at each assessment point (mean ± SD).

Dose (mg) BP (mmHg) Week 0 Week 3 Week 6 ∆ week 3 ∆ week 6

33

N=14

24hr 98.5 ± 8.2 97.8 ± 9.4 98.9 ± 9.4 -0.7 ± -3.4 0.4 ± 4.9 Night 85.8 ± 9.4 85.3 ± 10.5 87.4 ± 7.1 -0.5 ± -5.2 1.6 ± 4.9 Day 104.1 ± 7.5 103.5 ± 9.7 104.3 ± 10.5 -0.6 ± -4.5 0.2 ± 6.7

Clinic 107.1 ± 6.7 105.9 ± 8.2 106.5 ± 9.4 -1.2 ± -3.4 -0.6 ± -4.9

372

N=12

24hr 98.3 ± 8.0 98.4 ± 9.7 98.9 ± 9.4 0.1 ± 3.5 0.6 ± 4.5 Night 85.8 ± 8.3 85.8 ± 8.0 85.9 ± 8.7 0.0 ± 5.2 0.1 ± 6.0 Day 104.0 ± 8.7 104.0 ± 11.4 105.1 ± 10.7 0.0 ± 4.8 1.1 ± 5.2

Clinic 105.3 ± 6.9 104.6 ± 8.7 102.5 ± 8.7 -0.7 ± -5.5 -2.8 ± -4.1

712

N=13

24hr 95.4 ± 6.9 95.2 ± 7.2 96.0 ± 8.3 -0.2 ± -4.0 0.6 ± 3.6 Night 82.6 ± 7.6 84.1 ± 8.3 85.8 ± 9.7 1.5 ± 6.9 3.2 ± 4.7 Day 101.0 ± 6.5 99.9 ± 7.6 100.9 ± 7.9 -1.1 ± -4.0 -0.1 ± -4.0

Clinic 105.0 ± 8.3 103.0 ± 9.7 102.8 ± 8.7 -2.0 ± -5.4 -2.2 ± -5.8

1052

N=13

24hr* 93.8 ± 6.9 89.2 ± 8.7 90.8 ± 7.2 -4.6 ± -4.0 -3.0 ± -3.2 Night 82.8 ± 7.9 79.2 ± 7.6 80.8 ± 5.8 -3.6 ± -5.4 -2.0 ± -6.1 Day 97.9 ± 7.2 93.8 ± 9.7 95.2 ± 8.7 -4.1 ± -4.0 -2.7 ± -4.7

Clinic 103.1 ± 7.6 100.7 ± 9.7 100.4 ± 9.4 -2.4 ± -7.2 -2.7 ± -7.9

* Significant dose x time interaction (p=0.047)

24

Table 6: Heart Rate (beats/min) for 24 hour, night, day and seated clinic measurement in each

treatment group (mg flavanol/day) at each assessment point (mean ± SD).

Dose (mg) BP (mmHg) Week 0 Week 3 Week 6 ∆ week 3 ∆ week 6 33

N=14

24hr 68.1 ± 9.7 70.8 ± 10.5 70.6 ± 10.9 2.7 ± 5.2 2.5 ± 6.0 Night 60.4 ± 8.6 62.4 ± 8.2 61.6 ± 10.1 2.0 ± 3.4 1.2 ± 4.9 Day 71.5 ± 10.9 74.2 ± 12.0 74.4 ± 11.2 2.7 ± 6.4 2.9 ± 6.7

Clinic 64.6 ± 13.1 65.0 ± 13.1 63.9 ± 12.3 0.4 ± 4.9 -0.7 ± -6.4

372

N=12

24hr 72.7 ± 10.4 72.5 ± 9.7 73.5 ± 10.4 -0.2 ± -2.4 0.8 ± 5.9 Night 64.4 ± 10.7 64.3 ± 11.4 64.3 ± 9.7 -0.1 ± -3.8 -0.1 ± -5.9 Day 76.3 ± 10.0 76.3 ± 9.4 77.9 ± 10.7 0.0 ± -3.5 1.6 ± 5.2

Clinic 72.4 ± 14.2 70.0 ± 10.7 69.2 ± 11.1 -2.4 ± -6.6 -3.2 ± -6.6

712

N=13

24hr 68.6 ± 8.7 70.4 ± 10.5 69.7 ± 8.3 1.8 ± 4.0 1.1 ± 3.2 Night 61.2 ± 7.9 63.1 ± 10.8 61.9 ± 7.2 1.9 ± 4.7 0.7 ± 2.5 Day 71.9 ± 9.7 73.4 ± 10.8 73.2 ± 9.0 1.5 ± 4.3 1.3 ± 4.7

Clinic 64.8 ± 8.7 65.9 ± 9.0 64.6 ± 7.6 1.1 ± 6.1 -0.2 ± -4.3

1052

N=13

24hr 74.1 ± 5.4 74.3 ± 9.0 76.9 ± 9.4 0.2 ± 7.2 2.8 ± 7.2 Night 67.3 ± 6.5 66.3 ± 7.6 70.8 ± 7.2 -1.0 ± -5.0 3.5 ± 4.3 Day 77.0 ± 6.1 78.1 ± 10.5 79.8 ± 10.5 1.1 ± 9.4 2.8 ± 9.0

Clinic 74.3 ± 7.2 74.1 ± 11.9 71.8 ± 10.8 -0.2 ± -9.7 -2.5 ± -9.4

25

Table 7: Summary of study

What is known about this topic? What this study adds

• Consumption of cocoa rich chocolate has been shown to lower blood pressure (BP)

A meta-analysis of several intervention trials has demonstrated a BP lowering effect of cocoa rich dark chocolate compared to cocoa free white chocolate. • Cocoa flavanols (CF)improve endothelial

function Isolated CF and cocoa solids have been shown to enhance endothelial vasodilatation in the same manner as cocoa rich chocolate. • Lack evidence for effect of CF on BP Chronic consumption of isolated cocoa solids delivering CF doses in excess of those provided in chocolate have not lowered BP.

• Robust evaluation of effects of isolated

cocoa solids on BP Dose response evaluation of CF on ambulatory BP in borderline/mild hypertensives. • The study identifies a threshold dose for

cocoa flavanols Only the highest dose of CF (1052mg/day) lowered BP. This would account for lack of effect on BP with sub-threshold doses of CF in previous studies.

26

Figure 1: Consort diagram * withdrawn for personal or health reasons not related to study protocol;

** excluded due to gastric symptoms possibly related to cocoa beverage; ***excluded due to non-

compliance to study protocol (deliberate weight loss during intervention period).

27

Figure 2: Dose-related effects of cocoa flavanols on 24 hour ambulatory blood pressure.

Values represent means ± SEM of changes from baseline (average of changes to 3 and 6 weeks).

*Significantly different from all other doses by nested analysis (p<0.001).

28

Figure 3: Dose-related effects of cocoa flavanols on daytime and overnight ambulatory blood

pressures. Values represent means ± SEM of changes from baseline (average of changes to 3 and 6

weeks).